Upload
ngonhu
View
213
Download
1
Embed Size (px)
Citation preview
Modelling Invasive Processes in
Biology
Philip K Maini
Wolfson Centre for Mathematical Biology
Mathematical Inst
Oxford
UK
Outline
Epithelial sheets (rosettes)
Cranial neural crest cell invasion
Acid-mediated invasion hypothesis (cancer)
Shankar Srinivas Aaron Smith
Bradley Joyce David Kay
Abigail Moore Ruth Baker
Tristan Rodriguez
Jacintha Sugnaseelan
Georgios Trichas
Natalia White
Vivienne Wilkins
• The AVE then induces anterior pattern in the underlying epiblast by restricting expression of posterior markers to the opposite side of the epiblast cup
• In Nodal and Cripto mutants the AVE forms but does not migrate � severe gastrulation defects and failure to develop further
Rosettes form by cellular rearrangement:
• time-lapse (AVE cells in green, expressing Hex-GFP). Scale bar 50 microns
Junctional rearrangements
• We allow vertices closer than a certain
threshold distance to join together.
Vertex rearrangement
known as a T1 swap
Cell growth and mitosis
• Cells are assigned a certain volume and which
grows over time
• Cells can divide when they reach a certain
volume
Comparison with experiment: I
• Reduction in mean polygon number in the Epi-VE (pre- and post-migration) agreeing with observations:
Comparison with experiment: II
• Similar trends observed – 6-sided cells relatively less frequent in Epi-VE, while 4-sided cells more frequent there
Abnormal AVE migration in ROSA26 mutant
• Has disrupted PCP [Planar Cell Polarity] and forms fewer rosettes
Conclusions
• Our simple model qualitatively captures a lot of features observed in AVE migration
• Rosettes not essential for successful AVE migration
• Rosettes are essential for ordered migration observed in vivo
• Trichas, Smith, White, Wilkins, Watanabe, Moore, Joyce, Sugnaseelan, Rodriguez, Kay, Baker, Maini, Srinivas, Multi-cellular rosettes in the mouse visceral endoderm facilitate the ordered migration of anterior visceral endoderm cells, PLoS Biology, 2012, 10(2), e1001256
Cell-based modelling approaches
• Cell-centred
• Vertex Models (Oster, Honda, Davidson et etc)
• Potts Models (Glazier, Graner, Hogeweg)
• WHICH IS CORRECT??
• Chaste – Cancer, Heart And Soft Tissue Environment
• Modular
• Developed in a collaboration between Computer
Science, Oxford and CMB – lead developers David
Gavaghan, Joe Pitt-Francis, James Osborne and Alex
Fletcher
• Individual cell movement at the discrete scale modelled using nonlinear force laws can be described by nonlinear diffusioncoefficients on the continuum scale.
• Therefore, we can (a) relate different discrete models of cell behaviour; (b) derive discrete, inter-cell force laws from previously posed diffusion coefficients
• P.J. Murray, C.M. Edwards, M.J. Tindall, PKM, Classifying general linear force laws in cell-based models via the continuum limit (Phys Rev E, 2012)
• Comparing vertex-based models (Fletcher, Osborne, PKM, Gavaghan, submitted)
Migration of Cranial Neural Crest Cells
• Paul Kulesa, Rebecca McLennan, Katherine
Prather, Jason Morrison
[Stowers Institute, Kansas]
• Louise Dyson, Ruth Baker
• NC cell population is crucial for proper development of the heart and peripheral nervous systems
• Is the cellular origin of the highly aggressive cancers, melanoma and neuroblastoma
Model and manipulation
• Can a chemoattractant (VEGF) produced by the overying ectoderm be sufficient for robust invasion?
Model Prediction and Validation
• A single chemotactic gradient with a single cell type is not a feasible mechanism. There must be at least 2 cell types – one chemotactic, one not chemotactic.
• By FACS (flow cytometry analysis) and by LCM (laser capture microdissection) show significant differences in expression of 19 out of 84 genes.
• Leading NC cells have upregulated expression of cell guidance and navigation genes (cell guidance factor recptors [EphA4]; integrins (Itgb5); MMPs [MMP2]; cadherins [Cdh7]). Trailing cells have upregulatedexpression of cadherins distinct from leading NC cells.
Loss of Trailing Cells
• Disrupts migration (maybe the effects of pushing are lost)
• No change – leaders forge ahead• Affects movement of leaders through a
global change in chemoattractant profile
Lead Cells Transplanted Proximally
• They maintain their ordering and migrate• May overtake the host cells as, after all,
they are the leaders
C = cell density, N = cell number, d= fixed jump size
Variable jump size (normally distributed)
Exclusion principle
References
• McLennan, Dyson, Prather, Morrison, Baker, Maini, Kulesa, Multiscale mechanisms of cell migration during development: theory and experiment, Development, 139, 2935-2944 (2012)
• L. Dyson, P.K. Maini, R.E. Baker, Macroscopic limits of individual-based models for motile cell populations with volume exclusion, Phys. Rev. E. 86, 031903 (5 pages) (2012)
Acid-Mediated Invasion Hypothesis
• A bi-product of the glycolytic pathway is lactic acid – this lowers the extracellular pH so that it favours tumour cell proliferation AND it is toxic to normal cells.
• R.A. Gatenby and E.T. Gawslinski, A reaction-diffusion model of cancer invasion, Cancer Research, 56, 5745-5753 (1996)
• In the asymptotic limit (small parameter is tumour cell diffusion divided by lactic acid diffusion) this is Fisher’s eqn.
• Using asymptotics (some subtilies arise) --determine parameter space in which gaps occur
• McGillen, Martin, Robey, Gaffney, PKM, Riken Proceedings (Nishiura 60th
birthday) in press • McGillen, Martin, Gaffney, PKM (J. Math. Biol. to appear)
Results
• Previously unseen potentially dangerous elevation in blood pHe resulting from bicarbonate therapy in mice – confirmed by our in vivo expts
• Limited efficacy of bicarbonate in humans
• Martin, Robey, Gaffney, Gillies, Gatenby, PKM, Predicting the safety and efficacy of buffer therapy to raise tumour pHe: An intergrative modelling study, Brit. J. Cancer, 106,1280-1287 (2012)
Buffer therapy most effective
(a) in elderly patients with renal impairments
(b) in combination with proton production inhibitors (such as DCA), renal glomular filtration rate inhibitors (eg non-steroidal anti-inflammatory drugs and angiotensin-converting enzyme inhibitors) or
(c) with an alternative buffer reagent possessing an optimal pK of 7.1 – 7.2 (pK = -log(dissociation constant) – does not elevate (to dangerous levels) blood pH in the same way that bicarbonate can!!
Conclusions
(a) Used a vertex-based model for AVE migration on the mouse ectoderm to investigate the role of rosettes
(b) Used a hybrid model to study neuralcrest cell invasion
(c) Used a PDE model to study tumour cellinvasion
(d) Showed how many of these approaches lead to novel nonlinear diffusion type systems
AcknowledgementsShankar Srinivas Katherine Prather Ruth Baker Jessica McGillenBradley Joyce Jason Morrison David Kay Eamonn GaffneyAbigail Moore Ian Robey Aaron Smith Louise Dyson Tristan Rodriguez Bob Gatenby Joe Pitt-Francis Helen ByrneJacintha Sugnaseelan Bob Gillies James Osborne Philip MurrayGeorgios Trichas Katherine Prather David Gavaghan Marcus TindalNatalia White Paul Kulesa Gary Mirams Carina EdwardsVivienne Wilkins Rebecca McLennan Alex Fletcher Natasha Martin
• FundingBBRSC, EPSRC (IB, 2020 Science, DTCs), NIH-NCI